The Quick Answer
The right chart depends on what you want to show:
- Comparing categories? → Bar chart
- Showing parts of a whole? → Pie or donut chart (6 categories max)
- Tracking change over time? → Line chart
- Showing a relationship between two variables? → Scatter plot
- Showing how data is distributed? → Histogram
The most common mistake is using a pie chart when a bar chart would be clearer. If you're unsure, a bar chart is almost always a safe default.
Why Chart Choice Matters
A chart translates numbers into shapes and positions that the human visual system can process quickly. The right chart makes patterns obvious. The wrong chart hides them — or worse, creates patterns that aren't real.
Consider a dataset of quarterly revenue for five products. A pie chart makes it hard to compare slices of similar sizes. A bar chart makes the comparison instant because humans are much better at comparing lengths than angles.
Chart selection isn't about aesthetics. It's about cognitive efficiency: how quickly and accurately can someone extract the message from the visual?
The Six Common Chart Types
Bar Chart
What it shows: Comparisons between discrete categories.
Use when:
- Comparing quantities across groups (sales by region, survey responses by option)
- Ranking items from largest to smallest
- Showing frequency counts for categorical data
Strengths:
- Easy to read, even with many categories
- Precise comparison — the eye compares bar lengths accurately
- Works for both small and large datasets
Limitations:
- Not ideal for showing change over time (use a line chart)
- Can feel cluttered with more than 15–20 categories
Example: "How many units did each product sell last quarter?" — a bar chart with one bar per product instantly answers this.
Try it: Bar Chart Generator
Line Chart
What it shows: Trends and changes over a continuous sequence (usually time).
Use when:
- Tracking a metric over time (monthly revenue, daily temperature)
- Comparing trends across multiple series
- Showing rates of change (steep slope = fast change)
Strengths:
- The slope immediately communicates direction and speed of change
- Multiple lines can overlay without clutter (up to 5–6 lines)
- Handles many data points well (dozens to hundreds)
Limitations:
- Misleading if the x-axis isn't continuous or equally spaced
- Starting the y-axis at a non-zero value can exaggerate changes
- Not suitable for categorical comparisons
Example: "How has our website traffic changed over the past 12 months?" — a line chart shows the trend, seasonality, and any anomalies at a glance.
Try it: Line Chart Generator
Pie Chart
What it shows: How parts make up a whole (proportions that sum to 100%).
Use when:
- Showing the composition of a single total
- You have 2–6 categories
- One or two slices dominate and you want to emphasize that
Strengths:
- Immediately communicates "part of a whole"
- Familiar — almost everyone understands pie charts
- Effective when one segment is clearly dominant
Limitations:
- Hard to compare slices of similar size (is 23% bigger than 21%?)
- Doesn't work well with more than 6–7 segments
- Cannot show change over time
- Often outperformed by a simple bar chart
Rule of thumb: If you need to add percentage labels to every slice for the chart to be readable, a bar chart would probably be clearer.
Example: "What portion of our budget goes to each department?" — a pie chart works if there are 4–5 departments with meaningfully different allocations.
Try it: Pie Chart Generator
Donut Chart
What it shows: Same as a pie chart — parts of a whole — with a hollow center.
Use when:
- You want to show proportions but need the center space for a summary number (e.g., total, key metric)
- Dashboard design where space efficiency matters
Strengths:
- The center area can display a total, label, or icon
- Slightly easier to compare arc lengths than pie slice angles (the perimeter is the primary visual cue)
- Modern, clean appearance for dashboards
Limitations:
- Same readability issues as pie charts — avoid with many categories
- The hollow center reduces the visual area, making small slices harder to distinguish
When to use pie vs. donut: Use donut when you want to display a total or label in the center. Use pie when you want maximum visual area for the slices. In practice, the difference is mostly stylistic.
Try it: Chart Generator (supports donut mode)
Scatter Plot
What it shows: The relationship between two numerical variables.
Use when:
- Exploring whether two variables are correlated (e.g., advertising spend vs. revenue)
- Identifying clusters, outliers, or patterns in two-dimensional data
- Showing density — where data points concentrate
Strengths:
- Reveals correlations, clusters, and outliers that summary statistics miss
- Handles large datasets (hundreds or thousands of points)
- Each point preserves individual data values
Limitations:
- Requires two numerical variables (not categorical)
- Overplotting with very large datasets (points overlap)
- Correlation ≠ causation — the chart shows association, not cause
Example: "Is there a relationship between hours studied and exam score?" — a scatter plot shows whether higher study hours tend to correspond with higher scores, and how strong that relationship is.
Try it: Scatter Plot Generator
Histogram
What it shows: The distribution of a single numerical variable across ranges (bins).
Use when:
- Understanding the shape of your data (normal, skewed, bimodal)
- Checking for outliers
- Visualizing frequency distributions (test scores, response times, ages)
Strengths:
- Reveals distribution shape that summary statistics hide
- Identifies skewness, clusters, and gaps
- Works with continuous data of any size
Limitations:
- Bin size affects interpretation — too few bins hide detail, too many create noise
- Only shows one variable at a time
- Often confused with bar charts (histograms have no gaps between bars)
Example: "What is the distribution of customer ages?" — a histogram shows whether most customers are young, old, or spread evenly, and whether there are distinct age groups.
Try it: Histogram Generator
Chart Selection Decision Rules
These rules cover the most common scenarios. Start with the question you're trying to answer:
"How do categories compare?"
→ Bar chart. Horizontal bars if category names are long. Sorted by value for ranking.
"What are the proportions?"
→ Pie or donut chart if 2–6 categories. Stacked bar chart if more categories or if you need to compare proportions across groups.
"How has this changed over time?"
→ Line chart. Use multiple lines for comparing trends. Add markers for key events.
"Is there a relationship between X and Y?"
→ Scatter plot. Add a trend line if you want to show correlation direction.
"What does the distribution look like?"
→ Histogram. Experiment with bin counts — start with 10–20 bins and adjust.
"I need to show a single key number"
→ A large number display (sometimes called a "big number" or KPI card) is often better than a chart. Not everything needs to be visualized.
Common Mistakes
Using a pie chart for comparisons
Pie charts encode data as angles, which humans estimate poorly. If two slices are 24% and 27%, most people cannot tell which is larger without labels. A bar chart makes the 3% difference immediately visible.
Truncating the y-axis
Starting a bar chart y-axis at a value other than zero exaggerates differences. A bar showing 98 vs. 100 looks dramatically different if the axis starts at 95, but trivial if it starts at 0. Line charts have more flexibility here — context determines whether a zero baseline is needed.
Too many categories in a pie chart
Beyond 6–7 slices, pie charts become unreadable. If you have many categories, either group small ones into "Other" or switch to a bar chart.
Using 3D effects
3D charts distort perception. A 3D pie chart makes front slices appear larger than back slices of the same size. A 3D bar chart makes height comparison less precise. Always use 2D.
Ignoring color accessibility
Approximately 8% of men and 0.5% of women have some form of color vision deficiency. If your chart relies on color alone to distinguish categories, add labels, patterns, or direct annotations. Red-green combinations are the most problematic.
Dual y-axes
Charts with two y-axes (one on each side) are almost always misleading. The relationship between the two series depends entirely on how the axes are scaled, which the creator controls. Two separate charts placed side-by-side are clearer and more honest.
Plotting too many series on one chart
More than 5–6 lines on a line chart or 6–7 segments in a pie chart creates visual noise. If you need to show more, consider small multiples (repeating the same chart layout for each category).
Chart Type Comparison
| Question | Best Chart | Avoid |
|---|---|---|
| Compare values across categories | Bar chart | Pie chart (if >6 categories) |
| Show composition (parts of a whole) | Pie / Donut (≤6 parts) | Line chart |
| Track trends over time | Line chart | Pie chart |
| Show relationship between two variables | Scatter plot | Bar chart |
| Show data distribution | Histogram | Pie chart |
| Highlight a single value | Big number / KPI card | Any complex chart |
| Compare proportions across groups | Stacked / grouped bar | Multiple pie charts |
Design Principles That Apply to All Charts
- Title clearly states the takeaway. "Revenue by Region, Q4 2025" is better than "Revenue Chart."
- Label axes and units. Never leave axes unlabeled. Include units (dollars, kilograms, seconds).
- Use color purposefully. Highlight the data that matters. Use muted tones for context, strong tones for the main message.
- Remove chart junk. Gridlines, backgrounds, and decorations that don't convey data should be removed or minimized.
- Maintain consistent scales. If comparing multiple charts, use the same axis ranges.
- Source your data. A small note citing the data source adds credibility.
Frequently Asked Questions
When should I use a pie chart instead of a bar chart?
Use a pie chart only when you're showing parts of a whole (proportions summing to 100%), you have 6 or fewer categories, and at least one segment is distinctly different in size. In most other cases, a bar chart is clearer because humans compare lengths more accurately than angles.
What is the best chart for showing trends over time?
A line chart. The continuous line encodes the direction and rate of change — upward slopes mean increases, steeper slopes mean faster changes. Use markers for individual data points if you have few observations.
Can I combine chart types?
Yes, but carefully. A common combination is bars with a line overlay (e.g., monthly revenue as bars with a cumulative total as a line). Avoid dual y-axes, which are easily misinterpreted. Each visual element should serve a clear purpose.
How many data points is too many for a bar chart?
Bar charts work well up to about 15–20 categories. Beyond that, consider grouping small categories into "Other," filtering to the top N, or using a different layout like a horizontal bar chart sorted by value.
What is the difference between a histogram and a bar chart?
A histogram shows the distribution of a single continuous variable by grouping values into bins. Bars touch because the bins represent adjacent ranges. A bar chart shows values for discrete categories. Bars are separated because the categories are independent.
Is a donut chart better than a pie chart?
Neither is objectively "better." A donut chart provides space in the center for a label or total, which is useful on dashboards. A pie chart uses the full area, making slices slightly easier to compare. The readability issues (difficulty comparing similar-sized segments) apply to both.
Should the y-axis always start at zero?
For bar charts, yes — truncating the y-axis exaggerates differences and misleads viewers. For line charts, it depends: zero-baseline is safer for general audiences, but domain experts may benefit from a narrower range that reveals subtle trends.
How do I make charts accessible?
Don't rely on color alone to convey meaning. Use labels, patterns, or direct annotations. Test with a color blindness simulator. Provide a text description or data table as an alternative. Use sufficient contrast between chart elements and the background.
What file format should I save charts in?
PNG works for most purposes — presentations, reports, social media, and web. It supports transparency (useful for slides with colored backgrounds). SVG is better for web use where scaling matters. Avoid JPEG for charts because compression creates artifacts around sharp edges and text.
How do I choose colors for my chart?
Use a consistent palette with enough contrast between categories. Start with a tool-generated palette rather than picking colors manually. Avoid pure red and green together (color blindness). Use one highlight color for the most important data point and muted tones for the rest.
What chart should I use for survey results?
It depends on the question type. For multiple-choice questions (which flavor do you prefer?), use a horizontal bar chart sorted by response count. For rating scales (1–5 satisfaction), use a bar chart or stacked bar. For yes/no questions, a simple percentage with a donut chart works.
Can I use charts for small datasets?
Yes, but keep them simple. With only 3–4 data points, a table or a few large numbers may be more efficient than a chart. Charts become more valuable as datasets grow because they reveal patterns that tables hide.